Premature luteinization during GnRH antagonists cycles in IVF adversely affects pregnancy and implantation rates

Abstract

Objective: The effect of premature luteinization has always been controversial in IVF cycles. A retrospective analysis of our data from antagonists cycles showed that a serum progesterone >1.2 ng/ml was significantly associated with a worse outcome in terms of pregnancy and implantation rate. Therefore we wanted to clarify this conclusion in a prospective manner. Design: Prospective observational study. Materials/Methods: In October 2001 a prospective study was started comparing GnRH antagonists cycles in IVF. All patients had a serum estradiol, LH and progesterone determination on the day of hCG injection. Premature luteinization was defined as a progesterone level >1.2 ng/ml. Cycles with and without premature luteinization were compared for age, serum estradiol, progesterone and LH level, number of oocytes retrieved and embryos transferred. Pregnancy was defined as the presence of a gestational sac with fetal hearbeat. Implantation rate was defined as the number of fetal heartbeats over the number of embryos transfered. Statistical analysis was done using the Fisher exact test, Mann-Whitney test, logistic regression analysis and ROC curve as appropriate. A p value <0.05 was considered significant. Results: Fifty-two ART cycles were included. One embryo transfer was cancelled because of risk of OHSS and all zygotes were frozen. The incidence of premature luteinization, pregnancy and implantation rates are detailed in Table 1. Premature luteinization on hCG day ocurred in 24/52 cycles (46.2%). No difference was found in age, serum estradiol and LH level on hCG day, number of oocytes retrieved or in the number of embryos transferred between those cycles with and with-out premature luteinization group. A statistically significant difference was only found in the number of gestational sacs observed (0.28 vs 0.78 p = 0.0208). No LH surge (>10 U/L) was noted in any cycle on hCG day. Logistic regression analysis showed the progesterone level to be significantly correlated to cycle outcome ( −1.417 p = 0.026). The progesterone level of 1.2 ng/ml area under the ROC curve was 0.721 with a 80% sensitivity, 55% specificity, 54% positive predictive value and a 80.4% negative predictive value assuming a 40% pregnancy rate in the no premature luteinization group.Table 1Incidence of premature luteinization and cycle outcome during GnRH-antagonists IVF cycles.P level ng/mlprr (95% C.I.)1.2<1.2Cycles (%)24 (46.2%)28 (53.6%)Embryo Transfers2328Pregnancies515Pregnancy rate per transfer (%)21.7%53.6%0.0240.405 (0.173 to 0.948)Implantation rate (%)10.6%30.1%0.0060.351 (0.160 to 0.769) Open table in a new tab Conclusions: These preliminary data from this group of patients showed that premature luteinization during GnRH antagonists IVF cycles was a frequent event and adversely affected the pregnancy and implantation rate. Progesterone elevations were not related to serum LH levels and may reflect the mature granulosa cell response to exogenous FSH exposure. A progesterone level >1.2 ng/ml is the best cut-off criteria for an adverse outcome. Probably GnRH antagonists cycles could benefit from additional routine progesterone monitoring. Supported by: IVI Valencia